Nitrofurantoin (Macrodantin, Macrobid) is an antibiotic prescribed for prophylaxis for some women who have a temporal association between intercourse and urinary tract infection (UTI) recurrence. Multiple adverse reactions have been reported since the drug was introduced in the 1950s. The following case is an example of nitrofurantoin-induced fever associated with elevated liver enzymes, myalgias, and arthralgias.

Case Presentation
A 57-year-old woman presented to the hospital complaining of malaise, weakness, nausea, aches in her upper arms and hips, and itchy hands. She had a history of repeated hospitalizations because of recurrent fever, transaminitis, rigors, weakness, myalgias, and arthralgias. She reported intermittent fever spiking to 104°F in the past 6 weeks, with rigors lasting up to 7 days after the fever.

Physical examination was significant for right upper-quadrant tenderness over the liver, bilateral shoulder and hip tenderness, and tachycardia. Her history was also significant for depression, but she took no medications or herbal remedies. She worked in a medical environment, with possible exposures to contaminated surgical instrumentation.

Figure 1 shows data from her hospitalization for 2 of these episodes.

Laboratory findings were significant for cyclically elevated liver enzymes (Figure 2) that corresponded with her fever episodes. Her white blood cell count was low (2.9/μL), platelet count was high (363,000/μL), and C3 level was low. The mean corpuscular volume (MCV) was elevated at 106 fL (indicating macrocytosis); however, her baseline MCV was also elevated (102 fL). Over the course of her previous 2 admissions, a wide range of laboratory tests were negative or within normal range. Test results ruled out vegetations, a mass, malignancy, or any metastatic process.

A diagnosis of drug-induced fever was suspected, but the patient denied taking any medications before these episodes. With each admission, her symptoms resolved without intervention, and her liver enzymes slowly returned to baseline. At her last admission, the patient was again asked about any exposure that could have precipitated her recurrent fever and other symptoms. This time she discreetly mentioned that she now realized that all these episodes had occurred after she had had intercourse. Her husband realized that she had not told us about the nitrofurantoin her gynecologist had prescribed as prophylaxis for her recurrent postcoital UTI. She was told to stop taking the nitrofurantoin, and she has not had any recurrences since then.

Discussion
The term “goalpost fever” refers to a fever spike at hospital admission and again within 5 to 7 days, as in this patient.¹ Between the fever spikes, the temperature is normal. This pattern has been associated with malignancies and adverse drug reactions.

Fever of unknown origin
In adults, a fever of unknown origin (FUO) is defined as a temperature >38.3°C (100.9°F) lasting for more than 3 weeks, with no obvious source, despite a thorough medical evaluation.² Because no cause was found for her recurrent fever episodes, an FUO diagnosis was assumed.

The differential diagnosis of FUO includes infections, malignancies, and autoimmune conditions, among other causes. Evaluation begins with a thorough history, physical examination, and standard laboratory testing. Important components of the initial assessment include updated serologies, viral cultures, and radiologic studies, such as computed tomography or magnetic resonance imaging.²

Recurrent UTI
Recurrent UTI in an otherwise healthy woman is a worldwide problem. Each year, UTIs account for approximately 9.9 million physician office visits in the United States.³ This number likely underestimates the true prevalence, because UTI is not a reportable illness in the United States. Acute cystitis in women is generally an uncomplicated infection.

Recurrent UTI is common but can be controlled with long-term prophylaxis. The benefits of prophylaxis were recognized in the preantibiotic era (eg, the use of methenamine or a “ketogenic diet”). Cranberry juice, which produces hippuric acid in the urine and therefore exerts a weak antibacterial effect, has been used successfully to prevent recurrence of UTIs.⁴ Evidence shows that women who are susceptible to recurrent UTIs have consistent vaginal mucosa colonization by enteropathogenic bacteria.⁵ Studies have also shown the presence of transient bacteriuria in the immediate postcoital period.⁵

Prophylactic therapy
Prophylaxis should be considered in patients who have 3 or more separate UTI episodes annually. Most of the literature about UTI prophylaxis involves the use of trimethoprim/sulfamethoxazole (TMP/SMX; Bactrim, Septra), TMP (Proloprim) alone, or nitrofurantoin. In the past few years, the fluoroquinolones have been added as prophylactic options. UTI prophylaxis increases the interval between symptomatic episodes but does not decrease the risk for bacteriuria. TMP/SMX or fluoroquinolone prophylaxis decreases the rate of recovery in patients with gram-negative uropathogens from the fecal reservoir.6 Nitrofurantoin’s efficacy provides intermittent sterilization of the urine and does not alter the fecal flora. Therefore, it demonstrates the least resistance among all antibiotic medications, making it even more appealing as a prophylactic agent.

There are 3 strategies for antimicrobial prophylaxis for patients who suffer from recurrent UTI:

1. Continuous antimicrobial prophylaxis (suppressive therapy), with daily administration of a low-dose antibiotic (TMP 50 mg, TMP/SMX 40 mg/100 mg, or nitrofurantoin 100 mg).⁶ This well-tolerated modality is well-suited for patients with 3 or more UTIs annually.

2. Patient-initiated self-treatment, which may be effective among educated women with lower recurrence rates (ie, less than 2 episodes annually).⁷ This approach works well with a fluoroquinolone. Of the 7 available fluoroquinolones, the 3 most often used for genitourinary tract infections are ciprofloxacin (Cipro), levofloxacin (Levaquin), and gatifloxacin (Tequin).⁶

3. The third strategy involves postcoital antibiotic administration, which is appropriate for patients with a temporal association of intercourse and UTI recurrence. The use of a fluoroquinolone or TMP/SMX, taken twice daily for 1 day, is the treatment of choice.⁵ 

But the drugs of choice for UTI in fertile or pregnant women are penicillins, cephalosporins, or nitrofurantoin. (Fluoroquinolones may cause cartilage problems in the fetus, and TMP could potentially cause megaloblastic anemia in the fetus because of its antifolate effects.⁵)

Postcoital treatment can be used if bacteriuric episodes are infrequent. This is probably not the best option if the patient has 2 to 3 episodes weekly, because the therapy becomes chronic therapy and is associated with decreased adherence. 

Nitrofurantoin
Because UTI recurs in up to 5% of pregnancies, prophylactic treatment with nitrofurantoin or cephalexin is recommended for recurrent UTIs in pregnant women.⁷ Although nitrofurantoin’s safety in pregnancy has not been established (this drug is classified as a pregnancy category B), it is used frequently during pregnancy when other antibiotics are contraindicated. It should not be given to mothers who have glucose-6-phosphate dehydrogenase (G6PD) deficiency, because of hemolysis risk in the mother and fetus. This drug is excreted into breast milk in low concentrations.⁸ 

Nitrofurantoin is a synthetic antimicrobial agent that is useful in treating UTIs linked to Escherichia coli, Enterococcus species, Staphylococcus aureus, and certain strains of Klebsiella species and Enterobacter species. Pseudomonas and most Proteus strains are resistant to this medication. Because of its low renal parenchymal concentrations, nitrofurantoin is limited to treatment of infection in the lower urinary tract. Its nonspecific mechanism of action, inhibition of bacterial acetylcoenzyme, and subsequent disruption of carbohydrate metabolism create bactericidal conditions at high urine concentrations (attainable only with normal renal function) and bacteriostatic conditions with low concentrations.

Treatment or prophylaxis with nitrofurantoin in young, healthy patients is generally effective; older patients with an age-related decline in renal function do not respond as well. If a patient’s creatinine clearance is <40 mL/min, urine concentrations of the drug are inadequate, and the subsequent elevated blood levels increase the danger of toxicity.⁸

The most common adverse reactions to nitrofurantoin are nausea, vomiting, and diarrhea. Other possible reactions include peripheral neuropathy, pseudotumor cerebri, intrahepatic cholestasis, and hepatitis that is similar to chronic active hepatitis.⁸ Hypersensitivity reactions can involve the skin, blood, liver, or lungs and include fever and chills, blood dyscrasias, hemolysis (in cases of G6PD deficiency), and liver damage.⁶ 

Drug-induced hepatic reactions
Hepatic reactions to drugs, including antibiotics, are infrequent and involve a few cases per million treated patients annually.⁹ The incidence of hepatic injury caused by nitrofurantoin is low (estimated at 0.0003%).⁹ The difficulty in predicting the incidence of drug-induced liver disease is illustrated by the delayed impact of hepatotoxicity on new drugs: hepatotoxicity is the most frequent cause of postmarketing withdrawal of medications, despite the rigors of preclinical testing.⁹

In researching the etiology of our patient’s reaction to nitrofurantoin, we found many studies and case reports involving similar reactions. Possible mechanisms of nitrofurantoin-induced hepatic toxicity are either immunoallergic or secondary to metabolic idiosyncrasy. Direct toxicity is unlikely, since hepatic injury is not dose-dependent and is relatively rare and unpredictable.¹⁰ Several findings support an immunologic cause for the association with lupus erythematosus–like clinical presentation, autoantibodies against nuclear antigen and smooth muscle, and hypergammaglobulinemia.

A mechanism of hypersensitivity for liver injury with nitrofurantoin is also supported by reports that show a rechallenge with the drug in a patient 17 years after the initial hepatic injury.⁹ One case report indicated that low-dose nitrofurantoin consumed in milk from a nitrofurantoin-treated cow caused hepatic injury in a hypersensitive person.⁹

The diagnosis of immune-mediated drug-induced liver disease relies on the temporal association between the drug and the clinical syndrome, the rate of improvement after cessation of the drug (rapid in hepatitis, slow in cholestasis), and the presence of concurrent allergic manifestations (eg, rash, eosinophilia, fever). A positive rechallenge is the strongest clinical evidence for a cause-and-effect relationship, but rechallenge is rarely justified. Knowledge of the track record of the drug is useful, particularly if several drugs are involved. Ultimately, it is best to suspect a drug reaction in all hepatitis cases that have no definite cause and to discontinue use of the suspected agent.¹¹

No specific medical therapies are available for any of the antibiotic-associated hepatotoxicities; supportive care is required when the liver disease is severe or advanced. Liver transplantation may be indicated for irreversible liver failure. Reports of the use of corticosteroids to correct liver injury showed no clinical benefit.⁹ Educating patients about the symptoms of liver disease may be the best way to identify patients with clinically significant hepatotoxicity associated with antibiotic therapy.⁹ 

The prognosis of autoimmune hepatitis in the setting of nitrofurantoin exposure is generally good, with rapid improvement after cessation of the drug. In one series, 18 of 20 patients recovered after withdrawal of nitrofurantoin; the 2 patients who died had continued to take the medication after developing hepatic side effects.⁹

Conclusion
Clinical awareness is necessary for a prompt and correct diagnosis of adverse drug reactions. Patients should be counseled on the possible side effects of prophylactic medications. Physicians should inquire about potential signs and symptoms after prescribing prophylactic medications. If symptoms occur, the patient should immediately discontinue all medications, if possible, until an adverse drug reaction has been ruled out. It is recommended that patients bring all their medications to their physician so that discrepancies, interactions, and possible offending agents can be identified. If a drug reaction is identified, alternative therapies should be considered cautiously.

Acknowledgments
We wish to thank Dr Charles Ponte, Professor of Clinical Pharmacy and Family Medicine, Robert C. Byrd Health Sciences Center, Schools of Pharmacy and Medicine, West Virginia University, Morgantown, for his assistance in the review of the manuscript.

References
1. Murray HW, Mann JJ. Goalpost fever [letter]. Ann Intern Med. 1975; 83:84-85.

2. Roth AR, Basello GM. Approach to the adult patient with fever of unknown origin. Am Fam Physician. 2003; 68:2223-2228.

3. Schappert SM, Burt CW. Ambulatory care visits to physician offices, hospital outpatient departments, and emergency departments: United States, 2001–02. Vital and Health Statistics. Series 13, No 159. Hyattsville, Md: National Center for Health Statistics; February 2006.

4. Brumfitt W, Hamilton-Miller JM. Efficacy and safety profile of long-term nitrofurantoin in urinary infections: 18 years’ experience. J Antimicrob Chemother. 1998; 42:363-371.

5. McLaughlin SP, Carson CC. Urinary tract infections in women. Med Clin North Am. 2004; 88:417-429.

6. Matalon A, Yinnon AM. Fever and chills caused by nitrofurantoin taken after sexual intercourse. Isr J Med Sci. 1991; 27:17-18.

7. Bass PF III, Jarvis JA, Mitchell CK. Urinary tract infections. Prim Care. 2003; 30:41-61, v-vi.

8. Black DJ, Ellsworth A. Practical overview of antibiotics for family physicians. Clin Fam Pract. 2004; 6: 265-289.

9. Thiim M, Friedman LS. Hepatotoxicity of antibiotics and antifungals. Clin Liver Dis. 2003; 7: 381-399, vi-vii.

10. Stricker BH, Blok AP, Claas FH, et al. Hepatic injury associated with the use of nitrofurans: a clinicopathological study of 52 reported cases. Hepatology. 1988; 8:599-606.

11. Liu ZX, Kaplowitz N. Immune-mediated drug-induced liver disease. Clin Liver Dis. 2002; 6:755-774.